What if the Luddites weren’t just skilled journeymen bitter about the demise of their guild system? What if their warnings about automation already came true?

A piece of recent business news got me asking these questions about something I don’t yet have all the right words for. The story is that in its half-year result, Spark NZ said it was on track for cost savings of $40–60 million this financial year, in part through AI and automation. Cost savings means fewer employees (although it’s not clear how many will go).

What’s clear is that Spark sees AI as a cost-cutting technology, rather than a revenue-generator, which implies that Spark considers its people primarily to be costs. This is an interesting admission since AI incentivises Spark to get rid of as many of its humans as possible, which is why people are worried that automation will destroy more jobs than it creates.

However, automation has accelerated for a century and more people are in employment today than ever before. What’s going on here? Who is right? While the employment numbers don’t lie, they aren’t telling the full truth either. For example, what kind of jobs are being created parallel to automation? It turns out that’s a dangerous question to ask.

What I’m slowly starting to realise is that automation has destroyed our jobs, it’s just that most of us didn’t notice because we were too busy doing bullshit jobs instead.

The late anthropologist David Graeber nearly stumbled onto this phenomenon in his book called Bullshit Jobs but pretended he didn’t see anything and kept on going. His thesis was that an expanding percentage of the economy is essentially just busywork. He included hundreds of self-reported tales of the woeful “jobs” people really do. Graeber then advised that we should rethink what constitutes “paid labour” and introduce a Universal Basic Income (UBI) as a solution for a world with diminishing opportunities for humans.

But all his anecdotes suggested that we already live in a world with UBI, just not in the way Graeber thought about it.

He was right to point out that hundreds of millions of people arrive at their jobs every day to perform work that doesn’t matter in the slightest, even to the company that employs them, let alone to the wider society. This meaninglessness leaves the people who do these bullshit jobs in a state of unyielding despair. They feel, correctly, that they aren’t essential (witness the use of that term during Covid). And yet, they all get paid. Sometimes quite well.

I have a question: Where is the money that pays for these “jobs” coming from?

It’s not coming from the government. The government isn’t interested in directly paying people to work at private, for-profit companies. But the government also doesn’t want companies to focus on the truly productive workers while replacing the superfluous with robots.

After all, “What can we do with all these people” is the core policy question that directs all actions of every government anywhere on the planet across all of history. Everything flows from this. There seem to be three answers: People work, don’t work or work at a kind of halfway point between working and not working (bullshit jobs).

Since automation is steadily diminishing the number of people who are truly productive for companies, and letting people be out of work is simply unacceptable from a security perspective, the third option – soak them up with bullshit jobs – is the only real answer. The strange thing is that companies, not the government, are paying people to do these bullshit jobs. The government pays for welfare, but companies will pay for paper pushing.

This didn’t make sense to me, since there is no economic or financial incentive for for-profit companies to keep people employed doing bullshit tasks that have almost zero impact on the bottom line. Every time a new automation solution is implemented, pointless jobs are cut. But a few months or years later, new pointless jobs are invented, and the cycle begins again. What’s going on here?

At first, I thought this situation may be a sort of equilibrium, a compromise between multiple major social forces. No one invented this solution. No one asked private companies to create nonsense jobs to soak up surplus labour. It’s simply the result of relentless technological progress clashing with the imperative for governments to keep unemployment rates low.

But for-profit companies are meant to be, well, for-profit, right? Why would they care about the imperatives of the central government’s social policies? Companies have a fiduciary responsibility to maximise revenue for their shareholders, so it makes no sense that they would spend their precious profits on supporting nonsense jobs. The plot thickens.

If the government is concerned about social stability and jobs are a key method for ensuring stability, then it should be the government’s responsibility to either provide work programmes or do what Graeber suggested and pay people a UBI so they stay (mostly) out of trouble.

Except the thing is: people don’t appreciate what they get for free. Economists have known this for a long time. It was even suggested during Covid-19 that if the government were to charge a small fee for every vaccine, inoculation rates would rise rather than fall. This is a weird flaw in human psychology but it’s real. People only value things that come with a price.

This means that UBI wouldn’t work as a simple handout. And when I say “wouldn’t work”, I should clarify what free money is meant to incentivise. Under a UBI model, people are paid to live because our natural creativity is being unfairly strangled by wasting time earning money so that we can pay for basic living expenses. In one swoop, UBI would remove all the bullshit jobs and let people blossom, or so goes the theory.

However, I don’t think it’s controversial to say that about 70–80% of people aren’t creative at all. They couldn’t write a poem or draw a picture if you paid them $10,000 a day. The government is correct to be nervous about what a million bored people might do if they were raised above the first two layers of Maslow’s hierarchy of needs. And it wouldn’t be good – for the government.

All that matters is that the government gets what it wants (low unemployment) and the companies get what they want (better machines for extra profit). That’s the negotiation. The compromise is that businesses will use a portion of their extra funds to employ enough people to do bullshit jobs so that they aren’t a problem. It’s not ideal. Of course, the government should be paying a UBI. But it can’t because no one would accept the free money.

I suspect that this weirdness would be more obvious if New Zealand had a manufacturing base. But we don’t. Our economy is mainly “services”, which is great if you’re hoping to create plausible deniability that many, many Kiwis are essentially wasting their time on bullshit jobs just so the government can keep them out of trouble.

This compromise between industry and government reveals a deeper, more awkward reality about our modern world: there is no such thing as a private company. All business is subordinate to the state.

If a company follows even a single external regulation, it is no longer private. It is de facto a public company. It may be de jure private, but following government regulations means a business is not in full control of its future. There is no “free market” when government interferes at any level with regulations.

Here is the answer to my riddle: where does all the money come from to pay for the bullshit jobs?

The answer is that any money made by a business is already the state’s money. Don’t mistake me here. It’s not that a business is using money printed by the state. The issue is that business is part of the state since it cannot make its own sovereign rules; therefore the government is paying for these bullshit jobs through the proxy of the “private” company. The great illusion is that Spark is a private (listed) company. It’s not. Spark is part of the New Zealand state.

If Spark were a for-profit company, as it claims to be, then it should be automating workflows and firing humans at a much faster rate. And yet it doesn’t. In exchange for giving up full sovereignty and keeping its share of people employed doing bullshit jobs, companies like Spark are allowed to use automation to replace human labour and generate a little extra profit for their shareholders.

Sooner or later, the employees it fires today will be picked up by another company to do a different kind of bullshit job, shuffled around the economy like a travelling circus. I’ve seen this happen before and it will happen again. It’s corporate welfare, but not the kind of corporate welfare you’ve been trained to hate.

I’m starting to think the machines did steal our jobs because all we have left is this merry-go-round of bullshit jobs that stops us from asking, “Hey, maybe it doesn’t have to be this way?”

Nathan Smith is a former business journalist and columnist at the NBR. He also worked as the chief editor at the New Zealand Initiative policy think tank. He is now a freelance writer and copy editor.